Calibrating a RTCC the best way I can

[Moriambar]

Hello,
I usually have many projects with RTCCs, and I would like to properly calibrate them, in order to not have to correct them every time.
Currently I actually do not have the proper equipment to do so and the results are disastrous, especially after 6-12 months.
Now that I have a bit of budget I thought about buying a proper frequency counter, and I thought about this bk precision one, for example.

The problem is that I don't know how to read the accuracy figures for the instrument. For example, if I make my RTCC output a 1Hz square wave, is the counter capable of giving me 1ppm accuracy?

Also: are there any counters <1700€ which are better (not used one 'cause I don't trust them)

Thank you

[MosherIV]

When I did this for a comercial project (alarm clock), we used the counter to measure the xtal freq (32.768KHz).
The difference from 32.768 would tell us the correction needed.
Trying to work out deviation based on 1Hz will be very tedious!

You will find it very hard to achieve 1ppm. Try for something like 1second per month.

Sorry, I did not procure the counter so I do not know what is good or bad.

[Moriambar]

When I did this for a comercial project (alarm clock), we used the counter to measure the xtal freq (32.768KHz).
The difference from 32.768 would tell us the correction needed.
Trying to work out deviation based on 1Hz will be very tedious!

You will find it very hard to achieve 1ppm. Try for something like 1second per month.

Sorry, I did not procure the counter so I do not know what is good or bad.

Thanks for the insight. I think you're correct about being realistic, I just wanted to set some high goals, in order to settle to the max possible with my budget. 1s/month is great.
Nevertheless I don't know what counter to get.

[Ian.M]

Errrr......... 1ppm error is approx 2.6 times greater than one second per month. 

If you want to calibrate to that level of accuracy you'd need a reference oscillator with an order of magnitude better accuracy and stability.     The  BK precision frequency counter you linked is only 1ppm, so as-is you wouldn't have a hope of trimming close than 5 seconds per month.  However it does support using an external timebase, so if you also  get a 10MHz GPSDO to drive its rear panel external timebase input you could achieve the required accuracy.

However I strongly doubt that the RTCCs you are using have sufficient long term stability to be calibrated as you hope.

If you don't need to be able to measure frequencies above 100MHz,  a seven digit frequency counter built around a PIC18 MCU is reasonably simple, cheap to build and fairly easy to code.  With a little more thought, its resolution can be extended in software to an arbitrary number of digits, at the cost of requiring an order of magnitude increase in gate time for each digit.  Accuracy will be as good as its clock source.

[Berni]

Always use the extra square wave output pin on the RTC (probing the crystal makes it drift off). Most RTCs are configurable for something faster such as 512 or 1024 Hz on that pin. This makes it easier to measure the frequency. It should be perfectly sufficient to measure it with a frequency counter and adjust it until the counter reads spot on.

If you want to be really spot on you can get a 10MHz GPS reference oscillator and run that into the counter.

As for really getting it spot on you can run the RTC and the GPS reference both into 2 channels of a scope and trigger on one of them. This makes the other clock drift left or right across the screen. This lets you see really really tiny differences in freqency. But for a RTC that sort of accuracy doesn't matter.

[Moriambar]

Thanks Berni and Ian.M;
I get your point about long term stability etc.

Regarding the 10MHz oscillator I found some 10MHz GPSDO on eBay... but I don't know whether to trust them.
Can you suggest me something please?

[Ian.M]

Most of the cheap GPSDOs out of China seem to be based on BG7TBL's design.  Take a look at: https://www.eevblog.com/forum/testgear/bg7tbl-gpsdo-master-reference/

[Moriambar]

Most of the cheap GPSDOs out of China seem to be based on BG7TBL's design.  Take a look at: https://www.eevblog.com/forum/testgear/bg7tbl-gpsdo-master-reference/

Yep, those are the ones I found. They seem to do the job, according to the thread. thanks